Quality assessment system, quality assessment methodology, server, and software.

Abstract

DEPCT66 Quality assessment systems, methods for quality assessment, servers and programs capable of making assessments. The quality of the target item can be verified with high precision, namely, through the 10a quality judgment system. It consists of 302 teaching data sets that categorize images containing defects in objects. Numerous inspection targets are aligned with the brightness that characterizes the defect. As mentioned, this creates a large, categorized teaching dataset and repository 460b. For storing a large number of machine learning models, each built on a set of components. A large amount of teaching data is available through machine learning model construction services (80 models in total). Machine learning and a 460a camera unit for taking pictures of the target object. The 460°C inspection and judgment unit is used to assess the quality of the target items, with 14 items inspected each. As captured by the 460a camera unit, based on a learning model. Many machines are stored in storage unit 460b and, the unit selects the most suitable model 306. The results of the 460c decision unit will be evaluated, and a suitable machine learning model will be selected. The best of many machine learning models. -----------------------------------------------------------

Description

Quality determination system, quality determination method, server and program

[0001] The present invention relates to a quality determination system, a quality determination method, a server, and a program.

[0002] Patent Literature 1 describes a service provision system that provides services using machine learning based on artificial intelligence, and includes machine learning means that inputs learning data based on information sent from a user and generates a general model modeled by machine learning, personalization means that personalizes the general model to a model suitable for the user based on the information sent from the user, and service provision means that provides a personalized service to the user using the personalized model, and the information sent from the user is used for both the machine learning and the personalization.

[0003] Patent Literature 2 describes a method for visualizing the position of a sound source, which visualizes the position of an arbitrary sound source in real time by associating it with a real space. This visualization method detects one or more sounds, locates the positions of the respective sound sources, converts information about the sound sources, including at least the positions of the sound sources, into visible information, and displays the visual information superimposed on a real image of the area around the sound source in real time.

[0004] JP 2016-48417 A JP 2004-77277 A

[0005] An object of the present invention is to provide a quality determination system, a quality determination method, a server, and a program that can determine the quality of an object to be inspected with higher accuracy.

[0006] The invention described in claim 1 is a pass / fail judgment system comprising: a teaching data group generation unit that classifies images of a plurality of objects to be inspected, including defects, according to brightness that characterizes the defects, and generates a plurality of classified teaching data groups; a memory unit that stores a plurality of machine learning models constructed based on the plurality of teaching data groups by a machine learning model construction service that is provided as a cloud computing service and constructs machine learning models; a camera unit that captures images of the objects to be inspected; a judgment unit that judges the pass / fail of each of the objects to be inspected, imaged by the camera unit, based on the plurality of machine learning models stored in the memory unit; and an optimal model selection unit that evaluates the results of the judgment by the judgment unit and selects an optimal machine learning model from the plurality of machine learning models.

[0007] The invention described in claim 2 is a pass / fail judgment system comprising: a teaching data group generation unit that classifies images of a plurality of objects to be inspected, including defects, according to parameters that characterize the defects, and generates a plurality of classified teaching data groups; a memory unit that stores a plurality of machine learning models each constructed based on the plurality of teaching data groups; a camera unit that captures images of the objects to be inspected; a judgment unit that judges the pass / fail of each of the objects to be inspected, imaged by the camera unit, based on the plurality of machine learning models stored in the memory unit; and an optimal model selection unit that evaluates the results of the judgment by the judgment unit and selects an optimal machine learning model from the plurality of machine learning models.

[0008] The invention described in claim 3 is a pass / fail judgment system described in claim 1, in which the teaching data group generation unit performs a process P1 to determine a feature value representing the brightness that characterizes the defect for each of the images, and a process P2 to classify each of the multiple images according to the feature value and generate the multiple classified teaching data groups.

[0009] The invention described in claim 4 is a pass / fail judgment system described in claim 1, in which the teaching data group generation unit performs a process P1 in which, for each of the images, from among a plurality of reference values ​​that serve as standards for determining the brightness that characterizes the defect, a plurality of reference values ​​that represent the brightness that characterizes the defect are obtained as feature values, and a process P2 in which the same number of images as the number of feature values ​​are prepared, the images are classified according to the plurality of different feature values, and the classified plurality of teaching data groups are generated.

[0010] The invention described in claim 5 is a method for determining whether an object is good or bad using the system for determining whether an object is good or bad, the method including the steps of: the teaching data group generation unit generating the plurality of teaching data groups; the memory unit storing the plurality of machine learning models; the judgment unit judging whether the object is good or bad based on the plurality of machine learning models; the optimal model selection unit selecting an optimal machine learning model from the plurality of machine learning models; and the judgment unit judging whether the object is good or bad using the machine learning model selected by the optimal model selection unit.

[0011] The invention described in claim 6 is a server connected via a network to a cloud computing service that constructs a plurality of machine learning models for determining the pass / fail of a plurality of objects to be inspected based on a plurality of teaching data groups in which images containing defects of the objects are classified according to the brightness that characterizes the defects, and a judgment device that determines the pass / fail of the objects to be inspected based on the plurality of machine learning models, the server comprising: a teaching data group generation unit that generates the plurality of teaching data groups; and an optimal model selection unit that evaluates the pass / fail results determined by the judgment device and selects an optimal machine learning model from the plurality of machine learning models.

[0012] The invention described in claim 7 is a program that causes a computer connected via a network to a cloud computing service that constructs a plurality of machine learning models for determining the pass / fail of a plurality of inspected objects based on a plurality of teaching data groups in which images containing defects of the plurality of inspected objects are classified according to the brightness that characterizes the defects, and a judgment device that determines the pass / fail of the inspected objects based on the plurality of machine learning models, as a teaching data group generation means that generates the plurality of teaching data groups, and an optimal model selection means that evaluates the pass / fail results determined by the judgment device and selects an optimal machine learning model from the plurality of machine learning models.

[0013] The invention described in claim 8 is a program that causes a computer connected to a server having a teaching data group generation unit that generates a plurality of teaching data groups in which images containing defects of a plurality of objects to be inspected are classified according to the brightness that characterizes the defects, a cloud computing service that constructs a plurality of machine learning models that each determine the pass / fail of the objects to be inspected based on the plurality of teaching data groups, and an imaging unit that captures images of the objects to be inspected to function as a storage means that stores the plurality of machine learning models constructed by the cloud computing service and a judgment means that judges the pass / fail of the objects to be inspected each imaged by the imaging unit based on the plurality of machine learning models stored in the storage unit.

[0014] The invention described in claim 9 is a pass / fail determination system including: a visualization device having a detector for measuring a physical quantity that changes due to signs of a malfunction of a monitored object, and generating a plurality of visualized images that visualize the physical quantity; a teaching data group generation unit that classifies the visualized images including a plurality of previously acquired defects according to parameters that characterize the defects, and generates a plurality of classified teaching data groups; a storage unit that stores a plurality of machine learning models constructed based on the plurality of teaching data groups; an optimal model selection unit that selects an optimal model that is an optimal machine learning model from the plurality of machine learning models; and a determination unit that determines whether the operating status of the monitored object is pass / fail, based on the plurality of visualized images generated by the visualization device and the optimal model selected by the optimal model selection unit.

[0015] The invention described in claim 10 is a pass / fail determination system including: a sound source visualization device having a microphone for identifying a sound source generated from a monitored object, and generating a plurality of sound source visualization images that visualize the sound source; a teaching data group generation unit that classifies the sound source visualization images, which include a plurality of previously acquired defects, according to parameters that characterize the defects, and generates a plurality of classified teaching data groups; a storage unit that stores a plurality of machine learning models constructed based on the plurality of teaching data groups; an optimal model selection unit that selects an optimal model that will be an optimal machine learning model from the plurality of machine learning models; and a determination unit that determines whether the operating status of the monitored object is pass / fail, based on the plurality of sound source visualization images generated by the sound source visualization device and the optimal model selected by the optimal model selection unit.

[0016] The invention described in claim 11 is a pass / fail determination system including: a vibration visualization device having a vibration sensor for detecting vibrations generated from a monitored object, and generating a plurality of vibration visualization images that visualize the vibrations; a teaching data group generation unit that classifies the vibration visualization images, each including a plurality of previously acquired defects, according to parameters that characterize the defects, and generates a plurality of classified teaching data groups; a storage unit that stores a plurality of machine learning models constructed based on the plurality of teaching data groups; an optimal model selection unit that selects an optimal model that will be an optimal machine learning model from the plurality of machine learning models; and a determination unit that determines whether an operating state of the monitored object is pass / fail, based on the plurality of vibration visualization images generated by the vibration visualization device and the optimal model selected by the optimal model selection unit.

[0017] The invention described in claim 12 is a pass / fail determination method using the pass / fail determination system described in claim 9, including the steps of: the visualization device generating the visualized image; the teaching data group generation unit creating the plurality of teaching data groups; the memory unit storing the plurality of machine learning models; the optimal model selection unit selecting an optimal machine learning model from the plurality of machine learning models; and the determination unit determining whether the operating status of the monitored object is pass / fail using the machine learning model selected by the optimal model selection unit.

[0018] The invention described in claim 13 is a server connected via a network to a cloud computing service that constructs a plurality of machine learning models for determining whether the operating status of a monitored object is good or bad, based on a plurality of teaching data groups in which a plurality of visualized images, which visualize physical quantities that have changed due to signs of a malfunction of the monitored object, are classified according to parameters that characterize the malfunction, and a determination device that determines whether the monitored object is good or bad, based on the plurality of machine learning models, the server including: a teaching data group generation unit that generates the plurality of teaching data groups; and an optimal model selection unit that selects an optimal machine learning model from the plurality of machine learning models.

[0019] The invention described in claim 14 is a program that causes a computer connected via a network to a cloud computing service that builds a plurality of machine learning models for determining whether the operating status of a monitored object is good or bad, based on a plurality of teaching data groups in which a plurality of visualized images, which visualize physical quantities that have changed due to signs of a malfunction of the monitored object, are classified according to parameters that characterize the malfunction, and a determination device that determines whether the monitored object is good or bad based on the plurality of machine learning models, as a teaching data group generation means that generates the plurality of teaching data groups and an optimal model selection means that evaluates the good or bad results determined by the determination device and selects an optimal machine learning model from the plurality of machine learning models.

[0020] The invention described in claim 15 is a program that causes a computer connected to a cloud computing service that constructs a plurality of machine learning models for determining whether the operating status of a monitored object is good or bad, based on a plurality of teaching data groups in which a plurality of visualization images that visualize physical quantities that have changed due to signs of a malfunction in the monitored object are classified according to parameters that characterize the malfunction, and an imaging unit that captures images of the object to be inspected, to function as a storage means that stores the plurality of machine learning models constructed by the cloud computing service and a judgment means that judges whether the object to be inspected, each imaged by the imaging unit, is good or bad, based on the plurality of machine learning models stored in the storage unit.

[0021] According to the present invention, it is possible to provide a quality determination system, a quality determination method, a server, and a program that can determine the quality of an object to be inspected with higher accuracy.

[0022] 1 is a configuration diagram of a quality determination system according to a first embodiment of the present invention. FIG. 2 is an explanatory diagram of a teaching data creation device provided in the quality determination system. FIG. 3 is an explanatory diagram of a server provided in the quality determination system. FIG. 4 is an explanatory diagram of a teaching data group generated by the quality determination system and a plurality of machine learning models constructed using these teaching data groups. FIG. 5 is an explanatory diagram of an image including a defect in an object to be inspected. FIG. 6 is an explanatory diagram of a horizontal gray value profile graph schematically showing a portion of the scanning position. FIG. 7 is a horizontal gray value profile graph of image A schematically showing a portion of the scanning position. FIG. 8 is a horizontal gray value profile graph of image B schematically showing a portion of the scanning position. FIG. 9 is a horizontal gray value profile graph of image C schematically showing a portion of the scanning position. FIG. 10 is an explanatory diagram of preprocessing by the quality determination system. FIG. 11 is an explanatory diagram of a terminal and an imaging unit provided in the quality determination system. FIG. 12 is a flow diagram showing the operation of the quality determination system. FIG. 13 is a configuration diagram of a quality determination system according to a second embodiment of the present invention. FIG. 14 is an explanatory diagram of a terminal and an imaging unit provided in the quality determination system. FIG. 15 is a configuration diagram of a predictive maintenance system according to a third embodiment of the present invention. FIG. 16 is an explanatory diagram of a server provided in the predictive maintenance system. FIG. 17 is an explanatory diagram of a terminal and a sound source visualization device provided in the predictive maintenance system. 10 is a flowchart showing the operation of the predictive maintenance system according to a fourth embodiment of the present invention;

[0023] Next, embodiments of the present invention will be described with reference to the accompanying drawings to facilitate understanding of the present invention. Note that in the drawings, parts that are not relevant to the description may be omitted.

[0024] First Embodiment A quality determination system 10a (see FIG. 1) according to a first embodiment of the present invention can use machine learning to determine whether the appearance of an object 14, which is a product manufactured by a user, is good or bad. The quality of the appearance is determined based on the presence or absence of defects such as scratches or foreign matter. The object 14 is, for example, a part of a transportation machine such as an automobile or an aircraft, or food. However, the object 14 is not limited to these parts or food. A quality determination service using this quality determination system 10a is provided to users by a service provider.

[0025] 1, the quality determination system 10a includes a teaching data creation device 20, a server 30, and an inspection system 40a. The teaching data creation device 20, the server 30, and the inspection system 40a are connected to each other via the Internet N.

[0026] The teaching data creation device 20 is, for example, a personal computer. The teaching data creation device 20 is managed by a service provider or a user of the inspection system 40a, and includes a teaching data creation unit 202 as shown in FIG. 2. The teaching data creation unit (an example of teaching data creation means) 202 can acquire an image of the object to be inspected 14 from, for example, a camera 460 provided in the inspection system 40a, and create the image of the object to be inspected 14 as teaching data for constructing a machine learning model. The teaching data creation device 20 functions as teaching data creation means by a program executed by the teaching data creation device 20. The teaching data creation device 20 may also be a mobile terminal such as a smartphone.

[0027] The server 30 is managed by the service provider, and as shown in Fig. 3, has a teaching data group generation unit 302, an optimum model selection unit 306, and a management unit 308. The teaching data group generation unit (an example of teaching data group generation means) 302 preprocesses images IMG1, IMG2, IMG3, ... (teaching data groups TD) of the object to be inspected 14 including a plurality of different defects, which are created by the teaching data creation device 20, as shown in Fig. 4, to create a set TDg of preprocessed teaching data groups, i.e., teaching data groups TD1, TD2, TD3, TD4, ...

[0028] Here, when each image IMG1, IMG2, IMG3, ... of the inspection object 14 including the defect D centered at position (Xd, Yd) is captured by, for example, a 1.3 megapixel camera, the size of each image will be 1024 px vertically (in the Y-axis direction) and 1280 px horizontally (in the X-axis direction), as shown in FIG. 5 . This image is represented by a horizontal gray value profile graph as shown in FIG. 6 . Here, the horizontal gray value profile graph is a graph in which the pixels of the image are scanned in order and the brightness corresponding to each scanning position is represented, with the horizontal axis representing the scanning position and the vertical axis representing the brightness. The pixel scanning direction is, for example, from the upper left of the image to the right (positive direction of the X-axis), as shown by the arrow in FIG. 5 , and this is repeated downward (positive direction of the Y-axis) until all pixels are scanned.

[0029] The horizontal gray value profile graph shown in Figure 6 indicates that the brightness of defect D is brighter than that of non-defective portions. However, the brightness of defect D is not always brighter than that of non-defective portions, and it may be darker than that of non-defective portions. In this way, the brightness of pre-specified defect D has characteristics different from the brightness of non-defective portions, and defect D is distinguished from and characterized by brightness (an example of a parameter). Note that the horizontal gray value profile graph shown in Figure 6 only shows a portion of the scanning position (near defect D), and the omitted range is indicated by a dashed line. The same applies to Figures 7A to 7C, which will be described later.

[0030] The preprocessing performed by the teaching data group generation unit 302 includes the following processing P1 and processing P2. In processing P1, as shown in Fig. 6 , the teaching data group generation unit 302 compares the brightness of a pixel that characterizes a defect D, the position of which has been specified in advance, for each of a plurality of images IMG1, IMG2, IMG3, ... (teaching data group TD) of the object to be inspected 14 captured in advance by the teaching data creation device 20, with a plurality of thresholds TH1 to TH10 (examples of reference values ​​that serve as standards for determining a brightness range) that are different in size, and obtains a plurality of thresholds included in the range of brightness of the pixel that characterizes the defect D as feature values ​​that represent the brightness that characterizes the defect D. In process P2, the teaching data group generation unit 302 classifies each image IMG1, IMG2, IMG3, ... (teaching data group TD) into teaching data groups TD1, TD2, TD3, TD4, ... according to the multiple feature values ​​obtained in process P1. However, in this process, the teaching data group generation unit 302 does not classify one image into one of the teaching data groups TD1, TD2, TD3, TD4, ..., but rather copies the original image to prepare the same number of images as the number of feature values, and classifies the copies into teaching data groups TD1, TD2, TD3, TD4, ... corresponding to each threshold value. Therefore, through this preprocessing, teaching data groups TD1, TD2, TD3, TD4, ... are generated from the teaching data group TD, as shown in FIG. 4.

[0031] Next, a specific example of this preprocessing will be described based on images A, B, C, etc., which are examples of images IMG1, IMG2, IMG3 of the inspection object 14, respectively. Image A, which corresponds to image IMG1, includes defect D1 whose position has been specified in advance as shown in FIG. 7A, and defect D1 is characterized by a brightness range of 155 to 205. Image B, which corresponds to image IMG2, includes defect D2 whose position has been specified in advance as shown in FIG. 7B, and defect D2 is characterized by a brightness range of 165 to 215. Image C, which corresponds to image IMG3, includes defect D3 whose position has been specified in advance as shown in FIG. 7C, and defect D3 is characterized by a brightness range of 155 to 230.

[0032] In the above-mentioned process P1, first, a plurality of threshold values ​​(examples of reference values) 20, 40, 60, 80, 100, 120, 140, 160, 180, 200, 220, and 240 are set, and for image A (see FIG. 7A), three threshold values ​​160, 180, and 200, which fall within the brightness range 155 to 205 that characterizes defect D1, are obtained as feature values.

[0033] In the aforementioned process P2, image A is copied to prepare three images A, the same number as the number of feature values, and each image is classified into a teaching data group TD1 corresponding to the feature value 160, a teaching data group TD2 corresponding to the feature value 180, and a teaching data group TD3 corresponding to the feature value 200, as shown in Figure 8.

[0034] The above-described processes P1 and P2 are also performed on the remaining images B, C, etc. That is, for image B (see FIG. 7B ), two thresholds 180 and 200 within the brightness range 165 to 215 that characterize defect D2 are obtained as feature values, and image B is classified into a teaching data group TD2 corresponding to the feature value 180 and a teaching data group TD3 corresponding to the feature value 200 (see FIG. 8 ). For image C (see FIG. 7C ), four thresholds 160, 180, 200, and 220 within the brightness range 155 to 230 that characterize defect D3 are obtained as feature values, and image C is classified into a teaching data group TD1 corresponding to the feature value 160, a teaching data group TD2 corresponding to the feature value 180, a teaching data group TD3 corresponding to the feature value 200, and a teaching data group TD4 corresponding to the feature value 220 (see FIG. 8 ). Furthermore, the remaining images included in the teaching data group TD (images other than images A, B, and C) are classified into corresponding teaching data groups TD1, TD2, TD3, TD4, etc. according to the obtained feature values, and a set of teaching data groups TDg (see Figure 4) is generated.

[0035] These teaching data groups TD1, TD2, TD3, TD4, ... classified according to the brightness that characterizes the defects become teaching data groups for constructing machine learning models M1, M2, M3, M4, ... that determine the pass / fail of the inspection object 14, respectively.

[0036] In the preprocessing, the images IMG1, IMG2, IMG3, ... (teaching data groups TD) are not limited to being classified into teaching data groups TD1, TD2, TD3, TD4, ... based on brightness, but may be classified based on parameters other than brightness. Examples of parameters other than brightness include hue, saturation, and value that constitute the HSV color space, and R (Red), G (Green), and B (Blue) expressed as gradations in the RGB color model. In other words, any parameters may be used as long as they can characterize defective portions in distinction from non-defective portions. Furthermore, the preprocessing may include filtering to make defective portions more prominent than non-defective portions.

[0037] 1 and 4 is provided as a cloud computing service, and can build a trained machine learning model based on uploaded teaching data. This machine learning model building service 80 is, for example, Cloud AutoML Vision provided by the Google Cloud Platform (GCP).

[0038] The optimal model selection unit (an example of an optimal model selection means) 306 (see Figure 3) can evaluate the pass / fail judgment results of the inspection object 14 using multiple trained machine learning models constructed by the machine learning model construction service 80 and select the optimal machine learning model.

[0039] The management unit (an example of a management means) 308 can manage the status of the inspection system 40a used by the user. Specifically, the management unit 308 can record operation information relating to the operation status of the imaging unit 460 (see FIG. 1) included in the inspection system 40a. This operation information will be described later.

[0040] The server 30 functions as a teaching data group generating means, an optimum model selecting means, and a management means by means of a program executed within the server 30.

[0041] 1, the inspection system 40a includes a terminal 440, a programmable logic controller (PLC) 450, and an imaging unit (an example of a determination device) 460 that images the inspection object 14 and determines whether the inspection object 14 is good or bad. The terminal 440, the PLC 450, and the imaging unit 460 are connected to each other by wired communication or wireless communication.

[0042] The terminal 440 is managed by a user. The terminal 440 is, for example, a mobile terminal such as a personal computer or a smartphone, and may be a higher-level controller of the PLC 450. As shown in FIG. 9 , the terminal 440 has a machine learning model receiving unit 440a, a control unit 440b, and an operation status output unit 440c.

[0043] The machine learning model receiving unit (an example of a receiving means) 440a can download each machine learning model constructed by the machine learning model construction service 80. Note that each machine learning model is downloaded via secure communication.

[0044] The control unit (an example of a control means) 440 b can control the PLC 450 and the imaging unit 460 .

[0045] The operation status output unit 440c (an example of an operation status output means) can output operation information relating to the operation status of the imaging unit 460. This operation information is, for example, information on the time from when the imaging unit 460 starts capturing images to when it finishes capturing images. The operation information may also be information on the number of images captured by the imaging unit 460.

[0046] The terminal 440 functions as a receiving means, a control means, and an operation status output means by a program executed within the terminal 440. Furthermore, the machine learning model receiving unit 440a, the control unit 440b, and the operation status output unit 440c are not necessarily all included in one terminal 440, and each unit may exist separately on multiple terminals connected to each other. Furthermore, the terminal 440 may have a teaching data creation unit 202 instead of the teaching data creation device 20 shown in FIG. 2 .

[0047] The PLC 450 is a controller that is managed by the user and controls an inspection device 470 that inspects the object under inspection 14 as shown in FIG.

[0048] The imaging unit 460 (see FIG. 9 ) is managed by a user and can capture images of the inspection object 14. The imaging unit 460 can also determine the quality of each image of the inspection object 14 based on a plurality of machine learning models. The imaging unit 460 is, for example, a camera equipped with a GPU (Graphics Processing Unit). However, the imaging unit 460 may also be a mobile terminal with a camera, such as a smartphone. The imaging unit 460 has a camera unit 460 a, a storage unit 460 b, and a determination unit 460 c.

[0049] The camera unit 460a can capture an image of the object under inspection 14 and acquire image data. The storage unit 460b can store a plurality of machine learning models downloaded by the machine learning model receiving unit 440a. The determination unit 460c can determine whether or not the object under inspection 14 has a defect, i.e., whether or not the object under inspection 14 is pass / fail, based on the image data of the object under inspection 14 captured by the camera unit 460a and the machine learning models stored in the storage unit 460b. The determination unit 460c can perform arithmetic processing using the machine learning models at high speed and is configured, for example, by a GPU.

[0050] Next, the operation of the quality determination system 10a (a method for determining the quality of the object under test 14) will be described with reference to Fig. 10. The quality determination system 10a operates in accordance with the following steps S1 to S9. Of steps S1 to S9, steps S1 to S7 are operations that serve as preparatory steps required before the actual quality determination is made, and the subsequent steps S8 and S9 are operations for the actual quality determination in the inspection process before the shipment of parts, etc. Note that, if possible, the order of steps S1 to S7 may be reversed or they may be performed in parallel.

[0051] (Step S1) The teaching data creation unit 202 (see FIG. 2) of the teaching data creation device 20 (see FIG. 1) creates multiple image data of the object to be inspected 14, which constitutes the teaching data group TD shown in FIG. 4, based on image data of the object to be inspected 14 captured by the imaging unit 460. The teaching data group TD is a group of multiple image data of the object to be inspected 14 including a specified defect and multiple image data of the object to be inspected 14 not including the specified defect. Note that types of defects include, for example, scratches, voids, stains, and foreign matter contamination. However, the defects to be inspected differ depending on the object to be inspected 14. The created images of the object to be inspected 14 are stored in a storage means (not shown) and transmitted to the server 30 shown in FIG. 3. Instead of being created by the teaching data creation unit 202, the teaching data group TD may be created manually based on image data of the object to be inspected 14 prepared in advance.

[0052] (Step S2) As shown in FIG. 4 , the teaching data group generation unit 302 of the server 30 preprocesses multiple images (teaching data group TD) of the inspection object 14 generated by the teaching data creation device 20 to generate multiple teaching data groups TD1, TD2, TD3, TD4, ... classified according to feature values. In the preprocessing, for example, multiple brightness thresholds TH1 to TH10 (see FIG. 6 ) are used. Note that each threshold may be obtained by calculating the maximum brightness value of the defect D and dividing this maximum value. Thereafter, in response to an operation by the service provider or a user, the server 30 uploads the preprocessed teaching data groups TD1, TD2, TD3, TD4, ... to the machine learning model construction service 80. In this way, the preprocessing of the teaching data, which requires know-how, is performed by the service provider rather than the user, allowing the user to easily implement the machine learning model-based pass / fail judgment system 10a.

[0053] (Step S3) Based on the uploaded teaching data group, the machine learning model construction service 80 constructs machine learning models M1, M2, M3, M4, .... The accuracy of each of the constructed trained machine learning models M1, M2, M3, M4, ... is verified by the optimal model selection unit 306 (see FIG. 3 ) of the server 30. Note that if the accuracy is lower than a predetermined standard, the process returns to the previous step S2, and the teaching data group generation unit 302 performs preprocessing by further applying filtering, etc.

[0054] (Step S4) In response to a user operation, the machine learning model receiving unit 440a (see FIG. 9 ) of the terminal 440 downloads each trained machine learning model constructed by the machine learning model construction service 80. Each downloaded trained machine learning model is transmitted to the imaging unit 460 via the terminal 440.

[0055] (Step S5) The storage unit 460b of the imaging unit 460 stores each machine learning model downloaded by the machine learning model receiving unit 440a.

[0056] (Step S6) The control unit 440b of the terminal 440 controls the PLC 450 and the imaging unit 460 to experimentally determine the quality of the object 14 manufactured by the inspection device 470 (see FIG. 1 ). In detail, as a preparation stage for determining the quality of the object 14 when shipping it, the camera unit 460a (see FIG. 9) of the imaging unit 460 captures an image of the object 14, and the determination unit 460c inspects whether or not there is a defect in each of the object 14 based on the captured image and a plurality of machine learning models stored in the storage unit 460b, determining that the object 14 is a good product if there is no defect, and determining that the object 14 is a defective product if there is a defect.

[0057] (Step S7) The test determination results from the previous step S6 are transmitted from the imaging unit 60 to the server 30 via the terminal 440. The optimal model selection unit 306 of the server 30 shown in FIG. 3 evaluates the quality of each machine learning model based on the test determination results and selects the optimal machine learning model (hereinafter referred to as the "optimal model"). Information on the selected optimal model is transmitted from the server 30 to the terminal 440.

[0058] As described above, the steps up to step S7 are preparatory steps required before the actual quality determination is made. The next step S8 and subsequent steps are the steps for the actual quality determination in the inspection process before the shipment of parts, etc.

[0059] (Step S8) The control unit 440b (see FIG. 9) of the terminal 440 controls the PLC 450 and the imaging unit 460, and the imaging unit 60 judges the quality of the inspection object 14 transported on the conveyor of the inspection device 470 (see FIG. 1) based on the selected optimal model. In detail, the camera unit 460a (see FIG. 9) of the imaging unit 460 takes an image of the inspection object 14, and the judgment unit 460c inspects whether or not there is a defect in the inspection object 14 based on the taken image and the optimal model stored in the memory unit 460b, and judges the inspection object 14 as a good product if there is no defect, and judges the inspection object 14 as a defective product if there is a defect.

[0060] (Step S9) The operating status output unit 440c (see FIG. 9) of the terminal 440 transmits operating information relating to the operating status of the imaging unit 460 to the server 30. The transmitted operating information is stored in a storage unit (not shown) of the server 30, and the operating status of the quality determination system 10a is managed in an integrated manner by the server 30.

[0061] As described above, the quality determination system 10a according to the present embodiment determines the quality of the object 14 using an optimal machine learning model selected from the multiple constructed machine learning models, thereby obtaining a more accurate determination result. Note that, depending on the type of camera unit 460a, the quality determination system 10a can determine the quality of the condition of the object 14 other than the external appearance. For example, if the camera unit 460a is an infrared camera, it is also possible to determine the quality of the internal condition of the object 14.

[0062] Second Embodiment Next, a quality determination system 10b according to a second embodiment of the present invention will be described. Components having the same functions as those in the quality determination system 10a according to the first embodiment will be denoted by the same reference numerals, and detailed description thereof will be omitted. As shown in FIG. 11 , the quality determination system 10b includes a teaching data creation device 20, a server 30, and an inspection system 40b. The inspection system 40b includes a terminal 442, a programmable logic controller (PLC) 450, and an imaging unit 462 that images the object under test 14.

[0063] The terminal 442 (an example of a determination device) is, for example, a personal computer, a smartphone, an MR device for realizing MR (Mixed Reality), or an AR device for realizing AR (Augmented Reality). As shown in FIG. 12 , the terminal 442 has a machine learning model receiving unit (an example of a receiving means) 440a, a control unit (an example of a control means) 440b, an operation status output unit (an example of an operation status output means) 440c, a memory unit (an example of a memory means) 460b, and a determination unit (an example of a determination means) 460c, and can determine the pass / fail of each of the inspected objects based on multiple machine learning models. Note that the terminal 442 functions as a receiving means, a control means, an operation status output means, a memory means, and a determination means by a program executed within the terminal 442.

[0064] The imaging unit 462 has a camera unit 460a.

[0065] That is, in this quality determination system 10b, the terminal 442 includes the storage unit 460b and the determination unit 460c that the imaging unit 460 according to the first embodiment includes. The terminal 442 may include some of the machine learning model receiving unit 440a, the control unit 440b, the operating status output unit 440c, the storage unit 460b, and the determination unit 460c, and the PLC 450 may include the rest. That is, it is sufficient that the inspection system 40b as a whole includes the machine learning model receiving unit 440a, the control unit 440b, the operating status output unit 440c, the storage unit 460b, and the determination unit 460c. Furthermore, the machine learning model receiving unit 440a, the storage unit 460b, and the determination unit 460c may be included in the server 30 shown in FIG. 11 , rather than in the inspection system 40b.

[0066] 9 and 12, the only difference between the quality determination system 10b according to this embodiment and the quality determination system 10a according to the first embodiment is that the storage unit 460b and the determination unit 460c, which are provided in the imaging unit 460, are provided in the terminal 442. Therefore, the operation of the quality determination system 10b is substantially the same as the operation (steps S1 to S9) of the quality determination system 10a, and therefore a description thereof will be omitted.

[0067] Third Embodiment Next, a predictive maintenance system (an example of a quality determination system) 10c according to a third embodiment of the present invention will be described. Components having the same functions as those in the quality determination system 10b according to the second embodiment will be assigned the same reference numerals, and detailed descriptions thereof will be omitted.

[0068] The predictive maintenance system 10c according to this embodiment can determine whether the monitored object is operating normally by measuring the sound emitted by the monitored object, and can be applied to predictive maintenance that predicts malfunctions in the monitored object. The monitored object is, for example, a mechanical device, specifically a press. However, the monitored object is not limited to a press, as long as it is a device or equipment whose malfunction can be predicted by sound.

[0069] As shown in FIG. 13 , the predictive maintenance system includes a sound source visualization device (an example of a visualization device) 500, a teaching data creation device 20, a server 33, and a terminal 443.

[0070] The sound source visualization device 500 has a camera (not shown) that captures an image of the monitored object 600 and multiple microphones 502 for identifying a sound source generated from the monitored object 600, and can output a sound source visualization image in which the sound source is visualized by superimposing the sound intensity distribution on an actual image of the area around the sound source in real time. This sound intensity distribution is expressed as visualized information in the form of a heat map, with different colors corresponding to the magnitude of the sound pressure. The sound source visualization device 500 is sometimes called an acoustic camera.

[0071] As shown in FIG. 2, the teaching data creation device 20 has a teaching data creation unit 202, and can create an image as teaching data by importing a sound source visualization image.

[0072] The server 33 is managed by the service provider, and as shown in Fig. 14, has a teaching data group generation unit 302, a determination unit 334, an optimal model selection unit 306, and a management unit 308. As shown in Fig. 4, the teaching data group generation unit 302 (an example of a teaching data group generation means) preprocesses the teaching data group TD created by the teaching data creation device 20, and creates a set TDg of preprocessed teaching data groups, i.e., teaching data groups TD1, TD2, TD3, TD4, ...

[0073] The judgment unit (an example of a judgment means) 334 can virtually judge whether the operating status of each monitored object 600 is good or bad based on the teaching data group TD and multiple trained machine learning models constructed by the machine learning model construction service 80.

[0074] The optimum model selection unit (an example of optimum model selection means) 306 evaluates the determination result by the determination unit 334 and can select the optimum machine learning model.

[0075] The management unit (an example of management means) 308 can manage the status of the terminal 443 or the sound source visualization device 500. In particular, the management unit 308 can record operation information relating to the operation status of the terminal 443 or the sound source visualization device 500. Note that the server 33 functions as a teaching data group generation means, a determination means, an optimal model selection means, and a management means by a program executed inside the server 33.

[0076] 15, a terminal (an example of a determination device) 443 is connected to the sound source visualization device 500. The terminal 443 has a machine learning model receiving unit 440a, a control unit 443b, an operation status output unit 443c, a storage unit 443d, and a determination unit 443e, and can determine whether the monitored object 600 is good or bad based on the optimal model.

[0077] The machine learning model receiving unit (an example of a receiving means) 440a can receive the optimum model selected by the optimum model selecting unit 306 from the server 33. The optimum model is received via secure communication.

[0078] The control unit (an example of a control means) 443 b can control the PLC 450 and the sound source visualization device 500 .

[0079] The operation status output unit (an example of an operation status output means) 443c can output operation information related to the operation status of the terminal 443 or the sound source visualization device 500. This operation information is, for example, information on the time from when the sound source visualization device 500 starts capturing images to when it finishes capturing images. The operation information may also be information on the number of images output from the sound source visualization device 500.

[0080] The storage unit (an example of a storage means) 443d can store the optimum model received by the machine learning model receiving unit 400a.

[0081] The determination unit (an example of a determination means) 443e can determine whether the operating status of the monitored object 600 is good or bad, based on the multiple sound source visualization images output by the sound source visualization device 500 and the optimal model stored in the storage unit 443d. Note that the terminal 443 functions as a receiving means, a control means, an operating status output means, a storage means, and a determination means, depending on the program executed inside the terminal 443.

[0082] Next, the operation of the predictive maintenance system 10c (a method for determining whether the operating status of the monitored object 600 is good or bad) will be described with reference to Figure 16. The predictive maintenance system 10c operates in accordance with the following steps S3-1 to S3-9. Of steps S3-1 to S3-9, steps S3-1 to S3-7 are preparatory operations, and the subsequent steps S3-8 and S3-9 are operations for determining whether the operating status of the monitored object 600 is good or bad. Note that, if possible, steps S3-1 to S3-7 may be performed in a different order or in parallel.

[0083] (Step S3-1) The teaching data creation unit 202 (see FIG. 2) of the teaching data creation device 20 (see FIG. 13) imports the sound source visualization images generated by the sound source visualization device 500 as teaching data, and creates the teaching data group TD shown in FIG. 4. The multiple sound source visualization images (teaching data group TD) are stored in a storage means (not shown) and transmitted to the server 33 shown in FIG. 14. Instead of being created by the teaching data creation unit 202, the teaching data group TD may be created manually based on a sound source visualization image of the monitored target 600 that has been prepared in advance.

[0084] (Step S3-2) As shown in FIG. 4 , the teaching data group generation unit 302 of the server 33 preprocesses the sound source visualization image (teaching data group TD) generated by the teaching data creation device 20 to generate multiple teaching data groups TD1, TD2, TD3, TD4, ... classified according to feature values. Note that in the preprocessing, for example, multiple brightness thresholds TH1 to TH10 (see FIG. 6 ) are used. Note that each threshold may be obtained by calculating the maximum brightness value of the defect D and dividing this maximum value. Thereafter, in response to an operation by the service provider or the user, the server 33 uploads each of the preprocessed teaching data groups TD1, TD2, TD3, TD4, ... to the machine learning model construction service 80. In this way, the preprocessing of the teaching data, which requires know-how, is performed by the service provider rather than the user, allowing the user to easily introduce a predictive maintenance system using a machine learning model.

[0085] (Step S3-3) Based on the uploaded teaching data group, the machine learning model construction service 80 constructs machine learning models M1, M2, M3, M4, .... The accuracy of each of the constructed trained machine learning models is verified by the optimal model selection unit 306 (see FIG. 14 ) of the server 33. If the accuracy is poor, the process returns to the previous step S3-2, and the teaching data group generation unit 302 preprocesses the sound source visualization image (teaching data group TD) using a different method, for example by applying a different filter process.

[0086] (Step S3-4) In response to a user operation, the server 33 (see FIG. 14) downloads each trained machine learning model constructed by the machine learning model construction service 80. Each downloaded trained machine learning model is stored in a storage unit (not shown).

[0087] (Step S3-5) The determination unit 334 of the server 33 inputs the teaching data group TD to each machine learning model stored in a storage unit (not shown), and virtually determines whether the operating state of the monitored object 600 is good or bad.

[0088] (Step S3-6) The optimum model selection unit 306 evaluates the result of the judgment made by the judgment unit 334 in the previous step S3-6 as to whether the operating state of the monitored object 600 is good or bad, and selects an optimum model from among a plurality of machine learning models.

[0089] (Step S3-7) The optimal model selected by the optimal model selection unit 306 is transmitted from the server 33 to the terminal 442. The transmitted optimal model is received by the machine learning model receiving unit 440a (see FIG. 15) and stored in the storage unit 443d.

[0090] (Step S3-8) This step S3-8 is a step of monitoring the monitored object 600. The control unit 443b of the terminal 443 controls the PLC 450 to operate the monitored object 600. Meanwhile, the sound source visualization device 500 measures the sound generated from the monitored object 600 and outputs a sound source visualization image at a predetermined cycle. The terminal 442 determines whether the operating status of the monitored object 600 is good or bad based on the output sound source visualization image and the optimal model stored in the storage unit 443e.

[0091] (Step S3-9) The operation status output unit 443c of the terminal 440 transmits operation information related to the operation status of the sound source visualization device 500 to the server 33. The transmitted operation information is stored in a storage unit (not shown) of the server 33, and the operation status of the predictive maintenance system 10c is managed in an integrated manner by the server 33.

[0092] As described above, the predictive maintenance system 10c according to the present embodiment determines whether the operating state of the monitored object 600 is good or bad using an optimal machine learning model selected from the plurality of constructed machine learning models, thereby enabling predictive maintenance with higher accuracy. Note that the sound source visualization system may be replaced with any visualization device that has a detector for measuring physical quantities that change due to signs of a malfunction occurring in the monitored object 600 and is capable of generating a plurality of visualized images that visualize the physical quantities.

[0093] Fourth Embodiment Next, a predictive maintenance system 10d (an example of a quality determination system) according to a fourth embodiment of the present invention will be described. Components having the same functions as those in the predictive maintenance system 10c according to the third embodiment (see FIG. 13) will be assigned the same reference numerals, and detailed descriptions thereof will be omitted.

[0094] The predictive maintenance system according to this embodiment can determine whether the monitored object is operating normally by measuring the vibrations emitted by the monitored object, and can be applied to predictive maintenance that predicts malfunctions in the monitored object. The monitored object is, for example, a mechanical device, specifically a press or a conveying device. However, the monitored object can be any device or equipment whose malfunction can be predicted by vibrations.

[0095] 17 , the predictive maintenance system 10d includes a vibration visualization device 700 (an example of a visualization device), a teaching data creation device 20, a server 33, and a terminal 443 (an example of a determination device). The vibration visualization device 700 has a plurality of vibration sensors 702 for detecting vibrations generated from the monitored object 600, and can output a plurality of vibration visualization images that visualize the vibrations detected by the respective vibration sensors 702. The vibration visualization images are images that represent visualized information using different colors according to at least one of the magnitude and frequency of the vibration, for example.

[0096] Here, in the predictive maintenance system 10d, the vibration visualization device 700 and the vibration visualization image correspond to the sound source visualization device 500 and the sound source visualization image in the third embodiment, respectively.

[0097] By using such a predictive maintenance system 10d and performing the aforementioned operational steps S3-1 to S3-9 (see Figure 16), the judgment unit 443e (see Figure 15) of the terminal 443 can determine that an abnormality has occurred when a malfunction occurs in the monitored object 600.

[0098] As described above, the predictive maintenance system 10d according to the present embodiment determines whether the monitored object 600 is good or bad using an optimal machine learning model selected from the multiple machine learning models that have been constructed, thereby enabling predictive maintenance with higher accuracy. Note that the vibration visualization system may be replaced with any visualization device that has a detector for measuring physical quantities that change due to signs of a malfunction in the monitored object 600 and is capable of generating multiple visualized images that visualize the physical quantities.

[0099] Although the embodiments of the present invention have been described above, the present invention is not limited to the above-described embodiments, and all changes in conditions that do not depart from the gist of the present invention are within the scope of application of the present invention.

[0100] 10a, 10b Good / bad judgment system 10c, 10d Predictive maintenance system 14 Inspected object 20 Teaching data creation device 30 Server 40a, 40b Inspection system 450 PLC 80 Machine learning model construction service 202 Teaching data creation unit 302 Teaching data group generation unit 306 Optimum model selection unit 308 Management unit 334 Determination unit 440 Terminal 440a Machine learning model reception unit 440b Control unit 440c Operation status output unit 442, 443 Terminal 443b Control unit 443c Operation status output unit 443d Memory unit 443e Determination unit 460 Imaging unit 460a Camera unit 460b Memory unit 460c Determination unit 462 Imaging unit 470 Inspection device 500 Sound source visualization device 502 Microphone 600 Monitored object 700 Vibration visualization device 702 Vibration sensor N Internet

Claims

DEPCT661. The quality judgment system consists of a training dataset generation unit that categorizes images containing numerous defects of the target object according to the brightness characteristic of those defects and generates a large number of categorized training datasets; a storage unit for storing numerous machine learning models, each built on the basis of the large training datasets, using a machine learning model construction service offered as a cloud computing service; a camera unit for taking images of the target object; and a judgment unit for judging the quality of each target object photographed by the camera unit, based on the numerous machine learning models stored in the storage unit.The optimal model selection unit evaluates the decisions of the other unit and selects the most suitable machine learning model from a large number of machine learning models.

2. The quality judgment system consists of a training data generation unit that categorizes images containing numerous defects in the target object according to the brightness characteristic of those defects and generates a large number of categorized training data sets; a storage unit for storing the numerous machine learning models, each built on the basis of this large training data set; a camera unit for taking pictures of the target object; and a judgment unit for judging the quality of each target object photographed by the camera unit, based on the numerous machine learning models stored in the storage unit.

3. The Optimal Model Selector Unit evaluates the decisions of the other unit and selects the most suitable machine learning model from among many such machine learning models.

4. The Quality Judgment System specified in Claim 1, which is a quality judgment system, in which the training dataset generation unit about the images performs processing P1 to find the characteristic brightness value that characterizes the defect and, processes P2 which classifies the many such images into individual categories based on this characteristic value and generates a large such classified training dataset.

5. The Quality Judgment System specified in Claim 1, which is a quality judgment system, in which the training dataset generation unit about the images performs processing P1 to find the many reference values ​​that characterize the defect's brightness from among many such reference values ​​that serve as references to judge the defect's brightness as an individual characteristic value and,The P2 processor will provide multiple such images equal to the number of such feature values ​​and classify each such image according to the many different such feature values ​​and generate a large number of such classified training datasets.

5. The quality judgment method used is the quality judgment system specified in Claim 1, which consists of steps in which the training dataset generation unit generates the large number of such training datasets, and, the steps in which the storage unit stores the large number of such machine learning models, and, the steps in which the judgment unit judges the quality of each such inspection target based on the large number of such machine learning models, and, the steps in which the optimal model selector unit selects the machine learning model from among the large number of such machine learning models, and,The process involves the decision-making unit assessing the quality of the inspected target using a machine learning model selected by the optimal model selection unit.

6. A server connected to a cloud computing service constructs numerous machine learning models to determine the quality of each inspected target based on a large training dataset. This dataset contains images with numerous defects in the inspected target, categorized according to the brightness characteristic of those defects. The decision-making unit then processes these numerous machine learning models via a network, acting as a server containing units that generate the training datasets.The optimizer model selection unit evaluates the quality determined by the aforementioned judgment set and selects the most suitable machine learning model from among many such machine learning models.

7. The program provides a computer connected to a cloud computing service for constructing numerous machine learning models to determine the quality of each target object based on a large training dataset where images containing numerous defects of the target object are categorized according to the brightness that characterizes those defects, and, the judgment set for determining the quality of each target object based on the numerous such machine learning models via the network serves as the means to create the training dataset for constructing the numerous such training datasets and,The optimal model selection path evaluates the quality determined by the aforementioned set of judgers and selects the most suitable machine learning model from among many such machine learning models.

8. The program provides a computer connected to a server with a training dataset generation unit to create a large training dataset in which images containing numerous defects of the target object are categorized according to the brightness characteristic of such defects; and, a cloud computing service for constructing numerous machine learning models to determine the quality of each target object based on the large training dataset; and, an imaging unit for taking images of the target object, serving as a storage method for the numerous machine learning models created by the cloud computing service.The decision-making process for judging the quality of each target object photographed by the imaging unit is based on numerous machine learning models stored in the memory unit.

9. The quality judgment system consists of a set of abstraction generators with detectors to measure physical quantities that vary due to defect indicators in the monitored target object and generate numerous abstraction images from the visualization of these physical quantities; a training data unit that categorizes these abstraction images containing numerous pre-sought defects according to parameters characterizing those defects and generates a large number of categorized training datasets; and a memory unit for storing numerous machine learning models, each built on the basis of this large training dataset.The Optimizer Unit selects the best possible machine learning model from among many such machine learning models, and the Quality Judgment Unit judges the quality of the operating state of each monitored target based on the numerous abstract images generated by the abstraction generator and the optimal model selected by the Optimizer Unit.

10. The Quality Judgment System consists of a sound source abstraction generator with microphones to identify the sound source emanating from the monitored target and generate numerous abstract images of the sound source by visualizing it, and a training data generation unit that categorizes these abstract images of the sound source, including numerous pre-sought defects, in accordance with the parameters that characterize those defects and generates a large number of categorized training data sets.The storage unit holds numerous machine learning models, each built on the basis of a large training dataset; the optimal model selection unit chooses the best machine learning model from among these many models; and the judgment unit judges the quality of the operating state of each monitored target based on the numerous abstract images of sound sources created by the sound source abstraction unit and the optimal model selected by the optimal model selection unit.

11. The quality judgment system consists of a vibration abstraction unit with vibration sensors to detect vibrations originating from the monitored target and create numerous abstract images of vibrations obtained from visualization of such vibrations; and,The training dataset generation unit categorizes abstract images of such vibrations, incorporating numerous pre-sought defects that correspond to the parameters characterizing those defects, and generates a large number of categorized training datasets. The storage unit stores numerous such machine learning models, each constructed based on this large training dataset. Finally, the model selection unit selects the optimal machine learning model from among the many such models.The judgment unit for judging the quality of the operating state of each monitored target is based on the numerous vibration abstractions generated by the vibration abstraction imaging unit and the optimal model selected by the optimal model selection unit.

12. The quality judgment method used by the quality judgment system specified in claim 9 is a quality judgment method which includes the steps in which the vibration abstraction imaging unit generates the abstractions, and, the steps in which the training data unit generates numerous training data sets, and, the steps in which the storage unit stores numerous machine learning models, and, the steps in which the optimal model selection unit selects a machine learning model from among numerous machine learning models, and,The process by which the decision unit assesses the quality of the operating state of the monitored target using the machine learning model selected by the optimal model selector unit13. A server connected to a cloud computing service for constructing numerous machine learning models to assess the quality of the operating state of each monitored target based on a large training dataset containing numerous abstract images obtained from visualizing varying physical quantities caused by indicators of defects in the monitored target, and a set of decision-makers for assessing the quality of each monitored target based on these numerous machine learning models via a network of servers consisting of training dataset generators for creating such numerous training datasets and,The model selection unit selects the optimal machine learning model from among many such machine learning models.

14. The program provides a computer connected to a cloud computing service to construct many machine learning models to determine the quality of the operating state of each monitored target based on a large training dataset in which many abstract images obtained from visualization of physical quantities with variations caused by indicators of defects in the monitored target are categorized according to parameters characterizing such defects, and, the judgment unit for determining the quality of each monitored target based on many such machine learning models via the network serves as a means of constructing the training dataset for constructing many such training datasets and,The optimal model selection path evaluates the quality determined by the aforementioned set of judgers and selects the best machine learning model from among many such machine learning models.

15. The program provides a computer connected to a cloud computing service for constructing numerous machine learning models to determine the quality of the operating state of each monitored target based on a large training dataset in which many abstract images obtained from visualization of physical quantities with variations caused by indicators of defects in the monitored target are categorized according to parameters characterizing such defects, and, an imaging unit for taking images of the monitored target serves as a storage method for the numerous machine learning models created by the cloud computing service, and,The decision-making process for judging the quality of each target item being inspected, as photographed by the photographic unit, is based on numerous machine learning models stored in the memory unit.

Images